Lightnews — Scholar-powered news

Reposted by Lukáš Pekárek

Richard Sever @richardsever.bsky.social · 3d

AI in biology: "I read about some new computational method...get excited because it’s exactly what I need...I try the method...Most of the time things don’t work...I have seen this play out so many times my default assumption is nothing is going to work" blog.genesmindsmachines.com/p/we-still-c...

We still can’t predict much of anything in biology

Biology is hard. Yes, even for AI.

blog.genesmindsmachines.com

1 19 60

Lukáš Pekárek @pekarekl.bsky.social · 5d

That's actually when we used the "transient" foldable furniture the most 😜

1 2

Lukáš Pekárek @pekarekl.bsky.social · 5d

I like the analogy and I often use it too even though in a different context - you cannot organize your house/flat with just walls. You have a lot of sub-room organisational units in your flat which can be moved around or (dis)assembled when necessary... 🤔

1

Reposted by Lukáš Pekárek

Alexis Verger 🧬🧫🧪 @alexis-verger.cpesr.fr · 9d

AI-generated papers
AI-generated peer-review reports
AI-generated images
AI-grant writing
Algorithms to spot promising science

I'm very worried, we should all be very worried

1 2 16

Reposted by Lukáš Pekárek

samuel mehr @mehr.nz · 13d

here are some helpful steps for academics who are considering creating genAI images to use for lab branding:

1. don't

3 38 220

Reposted by Lukáš Pekárek

Vaclav Veverka @vaclav-veverka.bsky.social · 21d

How do flexible regions of histone chaperones team up to handle histones? Together with Fred Winston’s lab
@harvardmed.bsky.social, we reveal new insights in our study just out in Mol Cell. Hats off to James Warner and Vanda Lux @iocbprague.bsky.social for their key contributions! dlvr.it/TNB145

9 14

Reposted by Lukáš Pekárek

Matthew L. Kraushar @matthewkraushar.bsky.social · 23d

Rainer and I did our last experiment together in the Kraushar Lab @molgen.mpg.de on Sunday - like many such days since our postdocs. Rainer will soon lead a team in the Emmanuelle Charpentier Dept @MPUSP - stay tuned for more!

1 1 8

Reposted by Lukáš Pekárek

Lars Hubatsch @larshubatsch.bsky.social · 23d

Amazing!!! 🧪 The 2025 Ig Nobel prize goes to scientists from the @mpipks.bsky.social for discovering a bullet proof version to make the classic pasta alla cacio e pepe! Including a deep understanding of the pasta sauce in terms of phase separation physics 😃
Congrats @gbart.bsky.social et al!
Links👇

1 8 36

Reposted by Lukáš Pekárek

Julius Brennecke @juliusbrennecke.bsky.social · 24d

Love RNA biology?

Join us to explore the piRNA pathway with structural and genetic approaches (see 👇👇).

PhD student/postdoc position co-supervised by Clemens Plaschka & myself.

DM or email us if you’d like to know more!

@vbcscitraining.bsky.social @imbavienna.bsky.social @impvienna.bsky.social

Julius Brennecke @juliusbrennecke.bsky.social · 25d

PIWI clade Argonautes are essential for transposon silencing. Without them, animals are sterile due to massive transposon activity.

But how does piRNA-guided target interaction translate into silencing?

PhD student Júlia Portell Montserrat has an intriguing answer

www.cell.com/molecular-ce...

20 18

Reposted by Lukáš Pekárek

Jesse Veenvliet @jesseveenvliet.bsky.social · 28d

This is indeed a unique opportunity for @ec.europa.eu. But to ensure that „excellence thrives where it is enabled, protected & shared“ the % of funded proposals needs to stay the same (or actually increase). Otherwise it’s just a waste of talent.

European Commission @ec.europa.eu · Sep 11

The 2025 call for #MSCA Postdoctoral Fellowships drew 17,058 applications worldwide.

A 64.6% increase from 2024, it's the most popular call in over four decades.

Excellence thrives where it is enabled, protected and shared — and Europe provides that environment.

→ europa.eu/!jcnGgP

#ChooseEurope

1 3 12

Reposted by Lukáš Pekárek

Mark Rubin @markrubin.bsky.social · Sep 9

"Hostility, exclusion and academic bullying should be formally recognised as research integrity issues because they directly impact who gets to succeed and whose voices are silenced."

By @maddipow.bsky.social

Research integrity needs a kindness agenda or we will lose ECRs

Responding to early-career researchers’ honest questions with accusations of misconduct is a travesty of open science, says Madeleine Pownall

www.timeshighereducation.com

3 29 85

Reposted by Lukáš Pekárek

BK. Titanji @boghuma.bsky.social · Aug 26

Health misinformation and disinformation are the biggest and most under prioritized threats to public health in the 21st century. They are eroding decades of progress in the control and spread of diseases in real time. If we do not address this like the existential threat it is, we risk everything!

7 130 400

Reposted by Lukáš Pekárek

Jen Herman @jukebarosh.bsky.social · Aug 16

My Friday night hot take is that replication studies are largely a waste of time & $

If a result is worth following up on, it's better to address from a different angle & see if it holds water. Strong ideas withstand this

Conversely, there R infinite ways a replication can fail that teach you zip

3 9 43

Reposted by Lukáš Pekárek

Elise Cutts @elisecutts.bsky.social · Aug 13

"...if we deploy LLM assistants in decision-making roles... they will implicitly favor LLM-based AI agents and LLM-assisted humans over ordinary humans"

Translation: Use the product, or the product will discriminate against you.

Great

AI–AI bias: Large language models favor communications generated by large language models | PNAS

Are large language models (LLMs) biased in favor of communications produced by LLMs, leading to possible antihuman discrimination? Using a classica...

www.pnas.org

1 1

Reposted by Lukáš Pekárek

Joseph Howley @illdottore.bsky.social · Aug 14

You hear stories about what an academic household does to children but today at bedtime my precocious rising kindergartner said “father will you read me a story about university leaders who believe LLM-based chatbots have made language teaching irrelevant?” and fortunately I had just the thing

41 160 1.3K

Reposted by Lukáš Pekárek

TanentzapfLab @tanentzapflab.bsky.social · Aug 13

The goal of a PhD is not to learn some facts or read a few papers or learn a bunch of techniques. The goal of a PhD is to learn independence, problem solving, how to finish things you start, resilience, & gain the ability to adapt & think creatively. Learning these things is hard.

4 87 340

Reposted by Lukáš Pekárek

paveltomancak.bsky.social @paveltomancak.bsky.social · Aug 10

Looks like private funders, HHMI,CZI etc., are not going to “save” US research. Europe is setting up mechanisms to attract US researchers & it may work. But, common, target junior talent, don’t shower money on superstars who will be flying professors & return to US when the science winter passes.

9 40

Reposted by Lukáš Pekárek

Pookleblinky @pookleblinky.bsky.social · Aug 8

PhD level knowledge of geography right here. Who would not immediately want this in control of literally every aspect of their life

radams @radamssmash.bsky.social · Aug 8

My goto is to ask LLMs how many states have R in their name. They always fail. GPT 5 included Indiana, Illinois, and Texas in its list. It then asked me if I wanted an alphabetical highlighted map. Sure, why not.

A shitty map created by ChatGPT5. it purports to be US States with an R in their name, but it's wrong and dumb and worthless just like LLMs.

10 47 180

Reposted by Lukáš Pekárek

ℳatt @matttomic.bsky.social · Aug 6

It's incredible that all of the people who would shit on Wikipedia as a source because "lol anyone can edit that" are now 100% reliant on asking questions to the lying robot machine that no human has any oversight over

44 590 2.5K

Reposted by Lukáš Pekárek

Pavel @spavel.bsky.social · Aug 6

AI is a wonder technology that is going to change every industry, which is why
- people have to be forced to use it
- people have to be tricked into paying for it
- people who do use it on purpose have to hide it from their peers to avoid being mocked

9 120 380

Reposted by Lukáš Pekárek

Giosuè Baggio @giosuebaggio.bsky.social · Aug 5

Imagine this at scale, unchecked: how is this not a fundamental threat to science?

(Also, this is a perfect example of how irrational and ironic it is for large publishing companies to use “AI” to generate content that is of a much lower quality than the human-created content they already own.)

Iris van Rooij 💭 @irisvanrooij.bsky.social · Aug 4

Was looking into what Cognitive Scientists mean by ‘domain-general’ cognition, and then hit upon this. This is definitely NOT how we use that term.

Tip: read the last sentence to understand what went wrong here.

Sigh.

Domain-General Cognitive Ability
In subject area: Psychology
Domain-general cognitive abilities refer to cognitive skills that are not unique to humans but are present across various primate species and potentially other mammals and birds, indicating a general factor underlying cognitive performance that correlates with brain size and is associated with selective attention and working memory capacity.
AI generated definition based on: Trends in Cognitive Sciences, 2012

21 38

Reposted by Lukáš Pekárek

Chris H. Hill @chillzaa.bsky.social · Jul 19

New preprint alert! 🚨👇 I'm delighted to share the latest research from @jemmabetts.bsky.social
in our group: ‘A new protein-dependent riboswitch activates ribosomal frameshifting’. Huge congratulations to Jemma for the first manuscript of her PhD! ✨

www.biorxiv.org/content/10.1...

A new protein-dependent riboswitch activates ribosomal frameshifting

Programmed -1 ribosomal frameshifting (PRF) is a ubiquitous translational control mechanism in RNA viruses, allowing them to change the relative abundance of proteins encoded in different reading fram...

www.biorxiv.org

2 22 69

Reposted by Lukáš Pekárek

SeasonalPun @pixelfish.bsky.social · Jul 17

You know what emergent tech of the last decade actually works well, and was adopted by millions?

3D printing. It’s great! So many applications.

But also nobody is sneering at you for not having or utilizing 3D printing. Nobody is trying to sneak a 3D printer into your garage without your consent

100 1.5K 5K

Reposted by Lukáš Pekárek

Westermann Lab @westermannlab.bsky.social · Jul 11

Congrats to Gianluca Prezza (@gprezza.bsky.social) for reveiving the 2025 PhD Award from the Friends of the Helmholtz Centre for Infection Research (HZI)!👏 The award honours excellent PhD theses from the field of life sciences at HZI partner universities. His thesis focused Bacteroides' RNA biology.

Helmholtz-Zentrum für Infektionsforschung (HZI) @helmholtzhzi.bsky.social · Jul 11

Beim gestrigen Festakt wurden zudem die HZI-Promotionspreise an Wenchao Li @ciim-hannover.bsky.social und @gprezza.bsky.social aus der Gruppe von @westermannlab.bsky.social @helmholtz-hiri.bsky.social verliehen. Wir gratulieren den beiden zu ihren ausgezeichneten Doktorarbeiten!

3 10

Reposted by Lukáš Pekárek

James Saenz @jamessaenz.bsky.social · Jul 4

Hot off the press, our latest review of RNA-lipid interactions -> The secret life of RNA and lipids: "There is no life without RNA or lipids. But could there be life with only RNA and lipids?" @tcjea.bsky.social @bcube-tud.bsky.social be-tud.bsky.social

www.tandfonline.com/doi/full/10....

The secret life of RNA and lipids

There is no life without RNA or lipids. But could there be life with only RNA and lipids? The discovery that RNA can catalyse reactions in addition to encoding information opened new directions for...

www.tandfonline.com

21 58